Abstract We demonstrated that repression of the p16/pRb pathway in human mammary epithelial cells (HMEC), a condition common to stem cells and many tumor cells, induces several phenotypes critical to malignant transformation, including dynamic epigenetic remodeling resulting in the targeted methylation of selected CpG islands. Repression of p16INK4a activity in HMEC increases expression of the E2F1 transcription factor and, consequently, of several of its downstream targets such as the chromatin remodeling polycomb group (PcG) proteins, EZH2 and SUZ12. These PcG proteins localize to specific genomic sequences that become silenced for transcription and subsequently modified by DNA hypermethylation. Given the importance of stromal-epithelial interactions in differentiation and tumor progression, we hypothesized that cells in this epigenetically-plastic state could be programmed by signals from stromal cells within the microenvironment to acquire epigenetic changes that promote tumorigenesis. We developed an in vitro model system where epigenetically-plastic cells are exposed to factors that induce epithelial-to- mesenchymal transition (EMT), an important malignant program. Consistent with our hypothesis, the EMT was accompanied by expression changes, de novo methylation of a group of loci, including the E-cadherin promoter, and clinically relevant phenotypic changes such as increased motility. We wish to extend this powerful model system to investigate both intrinsic and extrinsic regulatory factors necessary for programming gene expression and epigenetic changes that are acquired during tumor progression. To address intrinsic control, we hypothesize that abrogation of different members of the p16INK4a/pRB pathway in HMEC will each generate a different spectrum of epigenetic changes, dependent on expression levels of chromatin remodeling proteins and sites to which they are targeted. To address extrinsic control, we hypothesize that epigenetically-plastic HMEC can be differentially programmed by signals from surrounding fibroblasts isolated from different breast tumor subtypes. We will test these hypotheses through Specific Aims that seek: (1) to determine if the program of expression changes that modulate levels of (a) chromatin remodeling proteins and (b) differentiation proteins that accompanies the loss of p16INK4a activity is phenocopied by expression changes for other members of the pRb pathway, i.e pRb, cyclin D1 and Cdk4, (2) to determine if the program of acquired DNA methylation events that accompanies the loss of p16INK4a activity is phenocopied by expression changes for other members of the pRb pathway, (3) to determine the effects of Luminal B and Basal-like carcinoma-associated fibroblasts on (a) gene expression changes and (b) acquired epigenetic alterations in epigenetically -plastic HMEC in vitro and to (c) test the functional consequences of these effects in vivo.